Involute Spline Size Inspection
by
Abstract
This article describes anew technique Ior
the' size determination of external Involute
splin ,by using a span measuring method.
It provides application performance informa-
tion demonstrating how this method and its
measurements correlate with the 'traditional
splin ring gage s~zing method.
Involute splines provide a positive
rotational coupling betweena shaft with
external teeth and related mating
member with internal spline teeth. Their
use permits ease of assembly or
disassembly Ier replacement or servicing,
and permits fixed or sliding connections.
Other uses permit compact assembly of
parts, or by use of standardized fittings,
the intereoupling of motors with gear
boxes or other loads.
Since there is no l10lling action between
the mating members. as there is between
meshing gears. aU of the spline teeth are
expected Ito fit together. AI.though
various internal andexter:nal fits are
available, the final goal of involute
splines isto achieve a self centering con-
dition with full contact bearing. The
result oJ this would be equalized load
sharing and stress on all the teeth.
Various manufacturing induced errors
diminish lhis idea] equalized condition up
to the point where a spline may fail to
properly engage its mating member.
Some typical factors which affect proper
mating are: tooth thickness, space width.
index errors. profile variations. lead
variation and out of roundness. Among
the manufacturing processes the one
most difficult to deal with is heat
treating. It is particularly troublesome if
the part is slender. 'thin walled or tubular.
because distortion becomes extremely
likely and also probably irregular.
To screen out parts which witl1 not
mate properly or not mate at all, it is
necessary to inspect the product pri'or to
the assembly operation. A very common
William L [anninck & John C. Nielsen
ITW lUitron
Illinois Tool Works Inc.
way of doing this is by the use of solid
tool steel gages. For external splines. the
gages are internal toothed rings called
r.
"GO" and 'NO GO (See Fig. 1).
The "GO" ring is a composite gage
having a full complement of teeth and is
used to inspect maximum material con-
ditions. minimizing the chance of fit in-
terference. In other words, if the part fits
the "GO" gage. there is a high probability
of assembling with its mate later. The
"GO" gage cannot measure the looseness
er size of'the part being gaged. It can
only reject those parts that will not enter
the "GO" gage.
. The "NO GO" gage does not have a
full compliment of teeth. but rather only
two diametrically opposed groups of
teeth. It is used to inspect minimum
material conditions, thereby, governing
the maximum looseness. If the spline part
is thin walled. or tubular. or heat treated.
it is desirable to Itry the 'NO GOI' gage
in at least two places 90 degrees apart.
An ova] part. for example. may enter the
'NO GO" gage at the low point, but not
the high. Ovality reduces the chances of
getting the full use of all the teeth in the
final assembly. At this point a few defini-
tions would be in order. (See Fig. 2).
Profile Variation: Deviation from
specified involute profile normal to the
flank surface with zere established at the
pitch line. Positive errors reduce clear-
ance or increase interference. Negative
errors increase clearance and reduce con-
tact area.
Index Variation: The difference be-
tween the actual and true spacing of any
two teeth, adjacent or not.
Effect.ive Tooth Thickness: The space
width of a perfect mating internal spline
for a fit without looseness or int'erierence.
Actual Tooth Thickness: The actual
measured tooth thickness on the pitch
line which is equivalent to the effecnve
tooth thickness. less width allowance for
profile and index errors.
The "GO" composite spline gage is
used to inspect the effective tooth
thickness down to the form diameter,
These gages must be very precise. since
they are used as the perfect mating
member, and any tolerances allowed
would steal. from the part tolerance.
The ~NO GO" sector ring gage con-
talning opposing sectors of about two
teeth is used to check the minimum
actual tooth thickness.
Using these gages requires a degree of
manual dexterity. The "GO" gage needs
to be tried in only one angular position.
but some skill is required to start 'the
AUTHORS:
MR. W. L. JANN]NCK is 'lechnical
Manager for tlcle ITW-lIIitron Division of
llIinois Tool Works Inc. He has beenirlvolved
in the desigJl and application of many types
of metal cutting tools specializin.g jn the art>Q
of tooling and gaging for the manufacturl:' of
gt>ar:s,splines and sprockets of both gf!!lerating
and fanning types. He was Chief Engineer and
Manager of Product Engineering from 1979
through 1984 for the Illinois Eclipse Division.
He has served on various committees of the
AGMA and MCTI and is Chainnan of the
AGMA Cutting Tool Committee. He also
sensed 011 the SAE-ANSI Inool'ute Sp1im Com-
mit:tee and the ASME-AN51 Committee orl
Power Transmission Chail'15and Sprock:ets. He
has written several articles on tooling subjecfs
and has spoken at various SAE, 5ME, A5QC
and AGMA meetings. He was educated' -at
Northwestern University.
MR. JOHN C. NIELSEN has been a staff
member of the lTW corporate machine
development group for 10 years. Subsequently
he became supervisor of thegroup. fie hClS
been respol'15ible for the design or project
management of a wide variety of eq,uipmfmt
relating to new product entry or productivity
improvement for fastener. cutting tool and
capacitor manufacture and undenJ)ater cable
handling, He currently specializes in iJ15pection
il'15trument development.
Prior to ITW. his design experience was .in
the development of new .Dusinessmachi11es for
Bell {:, Howell - Ditto Qnd electronic
cou.ntenneasures equipment for Northrop
(HalIicrafters) .
He holds a Bachelor of Science .degree in
Mechanical Engineering from Purdue Univer-
sity and is a member of the American Society
of Mechanical Engineers,
MarchjAprfll 19851.5
 MEASUREMENT BETWEEN PINS
(FOR SMALL DIAMETER SPLINES,
USE MEASU'REMENTS WITH
THREE PIN,S)

'i}
- f-- - --+--+

dI
1$
GO
~ CIRCULAR SPACE
WIDTH- REF

~~
~
M,AJ'OR ~ I

01,0.
MIN
-
FORM
Olt.

~ *
Nor GO

Fig. I

16 Gear Technologv
 CIRCULAF'I CIRCULAR
SPACE WIDTH TOOTH THICKNESS

MAX MAX
ACTUAL , NOT GO SECTOR PLUG GAGE EFFECTIVE GO COMPOSITE RING GAGE
OR MEASUREMENT BETwEEN
MIN MIN
-ACTUAL REFERIENCE DIMENSION EFFECTIVIE Nor GO COMPOSITE RIING
(MEASUREMENT BETWEEN
MAX PIN,SI
MAX
EFFE.CTlVE NOT GO COMPOSITE PLUG ACTUAL REFERHICE DIMENSION
GAGE (MEASUREMENT OVER IPIN,sl
MIN MIN
GO COMPOSITE PLUG GAGE ACTUAL NOT GO SECTOR RING GAGE
OR MEASUREMENT OilER
PINS

I NT,EFlNAl
-

CTUAL TOOTH' THICIKNESS

EACH SPACE IS .....$IC WIDTH. THE PUFECT eKTERN""l SPLINE WITH

BAfolC TOOTH TH1CXNfSS INTU.
fEtES AT x.

EFFECTIVE TOOTH THICKN'E.SS

EOUALS SPACE WIDTH OF
PERFECT INTERNAL SPLIN,IE

UAL SPACE wlDT H

THf PERF'ECT INTfRNAl SPLINE fITS IN ""NY POSITION,IF AU TEETH OF TH EXTERNAl.
+'-'--'--EHECT IvE SPACE WI DT H ARE NARROWED BY fHE AMOUNT OF THE INTERFERENCE.
EQUALS TOOTH THICKNESS OF
PERFECT EXTERNAL SPLINE EFFECT OF VARlATIONS ON EXTERNAL SPLlNlE

THE 'peuCT UT'ERN ...l SPLINE FITS IN ""NY POSITION, IF ALL SPAC,ES Ale WIDENED
BY TH E""""O UNT OF Tli E INTEAFU[NCE.

EFFECT OF VARIATIONS ON INTERNAL SPUNE

Fig. 2

March/April 1985 17

MAJOR
D'IA
CIRCULAR TOOTH
THICKNESS
'X ...fr. tz-
FIg. J
spline into the gage, especially when the
fit betwee.n part and gage is close. If the
spline won't enter ful1y, it is classed as an
oversize reject. The "NO co. because
of its limited teeth and usual non entry
situation, is more difficult to apply.
Along with the difficuJty in use, industry
standards for this check are not consis-
tent for either the number of places,
which constitute a reject, nor for the
force required to attain engagement. It
is obvious that the use of "GO" and "NO
GO" gages is tedious and labor and skill
intensive ..
Another element necessary to consider
is gage weal'. Allowances an! made for
this in making a new gage. Periodically
active gages must be monitored for size.
This can be done with measurement be-
tween pins, but usually a tapered master
plug is used. (See iFig. 3). Etched lines
along the plug indicate the fit or wear
range, This means splines accepted by a
worn gage may be rejected by a new one.
Ring gage tolerances an! planned to
absorb no more than 25 % of the product
tolerance .. In practice, byway of a
sampling ofactual gages in active use, it
was found that the measured accuracy
18 'GearTecMology
10------ SPU N E
DEPENDENT
SIZE
LEAD
LENGTH- RANGE
RE'F RING
FIT
GAGE
RAN,G,E
wlTHllN
DETERMINED
of "GO" gages can absorb a higher
percentage.
For many years, the span or block
measurement system for sizing involute
gears and splines has been used. It is a.
simple procedure using a micrometer and
reads over 3. group of teeth. (See Fig A) .
[t is independent from the outside
diameter variations or runout. A normal
line to any tangent on an involute will
lie in a plane tangent to the base cylinder.
Therefore, two parallel span flanges will
contact opposing tooth profiles similar to
the measurement of a cylinder. This
means the micrometer can be rocked,
reading from base diameter to tip for a
constant reading, in either direction. As
in measuring a cylinder, the check is in-
dependent of runout.
A dynamic means for span measure-
ment has been developed to duplicate the
principle of 3. flange micrometer on block
measurement. This requires a support for
the piece part on its journals in a roller
cradle or between centers. The work
piece is approached by a rotating spin-
dle which is pivoted for access ..The spin-
dle carries a pair of helical checking discs
which engage the work piece over anap-
LENGTH -------1
ON OVER-
ALLOWANICE ----
W!EAR
RANGE
y z
xx ,YY AND ZZ
BY CALI,BRATION
propriate span of teeth as shown in Fig.
S. One disc is fixed axially on the drive
spindle and is keyed to rotate the spline
on its axis, The second disc is also keyed
to the drive spindle, but is free to slide
axially on the spindle. The sliding disc
is directed by a spring to close over the
span of spline teeth .. As the work piece
is rotated through one full revoletion, the
axial movement of the second disc senses
the variation in spans progressively tooth
by tooth. A displacement transducer
records the axial. movement of the sens-
ing disc and provides a continuous
signal, which includes a succession of
span readings equal to the number of
teeth in the part.
The number of teeth in the span for
30 degree PA splines falls in approx-
imat.ely a 60 degree sector, or for a 40
tooth part the span is 7 teeth. In the
check of this part. 40, 7 tooth span
readings will occur.
The output of the displacement
transducer may be displayed on an
oscilloscope, recorded on a strip chart or
made to trigger indicator lights. (See Fig.
6). This shows a strip chart trace for span
measurement of a typical spline. Each
 I n
UI
r -I I

JEI~Automatic IOear H'obb'ng

Machines, available in2 si:zes:
MODEL JDP2: MODEL JDp3 I

26" Gear Diameter 32 Gear Diameter

II I

D,P.4 D.P.3
I $46,500 $64,200

f Differential, Automatic two-cui. Ihobbin g cycle,

Hydraulic support lift . .HydrauliCcrowning attachmsnt,
Automatic hob shifting.

fig. 4

L AX~L TFIAVEL

PICCO Unive";l Chamferingl and

Gear ,Rounding Machl'ne
TIRANSDUCER MAX. MODULIE 15
MAX. DIA. 24"
MODEL K15, $47,800
DRIVEN SUiDING DISC

---wr-- SPAN
SPLINE

Fig. 5

- 0- -' I

H iBILOIN High S;peed :Small

amplitude excursion from one peak to To relate the span measurement signal
'Gear Ho'b'blng! Machine
120-12 ~O.P.
the next represents one span measure- to the "GO" - "NO GO" acceptance Dia. 0.118 - 6.299 II

ment. The 4S teeth in the spline ar,e limits, the following paramaters have Prices start at $44,200
represented by 4.5 peak to peak span been established. To determine an over-
measurements over 8 teet.h. One revolu-

tion of the part is obvious as the cycle
repeats again after 45 peaks. The
sinusoidal record for this particular spline
is an indication of piece part ovaliry
caused by heat treat distortion of a thin
waned, tubular part.
*
size condition, the maximum span
reading plus the span variation in one
revolution of the part are compared to
established preset limits. For determina- ILUES. AMERICA"INC"
5420 Newport Drive
tion of an undersized spline the max- Rolling Meadows, ,ll160008
imum span is compared to an established
3 1i2-577-1144
lower limit. In addition, span variation
CI:RCLIE A15 ON READER REPLYCARD
March/April 1985, 19'
is compared to an established limit, The
TYPICAL SPAN MEASUREMENT TRACES
size category determination, that is com-
parisian of the measured span to. limits
is accomplished in an electronic logic
package accompanying the mechanical
inspection unit, The package incor-
porates digital displays of the greatest
span measured and span variation as well
as indicator lights which classify the in-
spected spline as OVERSIZE, GCXJD, or
UNDERSIZE. Setting masters are used to
calibrate the instrument, PRODUCTION PARTS
GOOD UNDERSIZE
In the develapment of this span
measurement, three specific production
splined shafts were studied thoroughly,
(See Fig. 7). They were: turbine shaft,
sun gear shaft and oil pump shaft. The
turbine and oil. pump were rack rolled,
the sun shaft was broached.
With the assistance of an independent
quality laboratory, samples of the three
production shafts and their ring gages for
inspection were accurately measured for
various elements. The results of some of
the most important measured piece part
elements are shown in TABLE J. The
results ofcorresponding measurements
made on the "CO~composite ring gages
for these piece parts are shown in TABLE
II.
The tabular information indicates the
variability within the parts and the gages.
From the data it can be seen that a large
portion of the fit tolerance was used by
the ring gage itself. In the case of the oil
pump shaft, the accuracy of the produc-
tion piece parts was as good or better
than the inspection ring gage.
A correlation study of span measure-
ment to determine spline size, using ring TABLE 1
gages, for the 40 tooth turbine shaft is Measurement of Splined Product
shown in Fig. 8. It reveals a rather good
correlation, and in this case, the gray Element Turbine Sun Oil Pump
area is only a few tenth thousandths Range in inches
inches wide. 1) Total Index Variation .001 - .0029 .0036 - .01158 .000810- .001670.
The span measurement system pro-
2) Pitch Variation .00034 - .00087 .00074- ,00190 .000'290- .00D830
vides a measurement for size determina-
3) Spacing Variation .00032- .00111 .00063- .00197 .000300- ,001090
tion; whereas, the ring gages give a
somewhat subjective size determination.
In practice, it has been found that from
TABLE 2
the measurement resulting from the span Measurement of Ring Gages
system, ring gage Ht predictions can be ("GO" Composites)
made.
Splines determined to be uGOOD" by iElemenl Turbine Sun Oil Pump
the span system, and new, unworn ring Range in inches
gages may be "REJECT', as determined 1) Total Index Variation .00136 - .00162 .00175 - .00215
by worn, but still within tolerance ring
2) Pitch Variation .00021 - ,00025 .00075 - .00100 ,00037- ..00040
gages. This disparity can be troublesome
3) Spacing Variation .00024 - .00037 .00143 - .00176 .00029 - .00038
to a supplier of splines and a user of

Mcrch/AprUI985 21
 splines both checking parts with ring
gages, but one using a worn set and one
usinga. new set. The span system
precludes this possibility.
Furthera.pplication of the span
measurement system to a. broader range
of products willeventuall'y determine the
limit of this process. In thefuture, ring
gages may be merely a periodic check of
size correlation much as tapered plug
gages are a periodic check of ring gages
for wear today.

Portions of this paper have been presented at the Fall

1984 Technical Conferences for AGMA & SME.

E-2 ON READER REPLYCARD

fig. '1

.5775
Fig. 8 Turbine Shaft - 40 Tooth. Spline Correlation
o
o
..5770 0 e e e OVERSIZE
e e o

.5765
0
0
0
0
e
~
0

~
rI
~ ~ ~ o e ~
~ (] ~ m e GJ Q
~ t:J
.5760 C!l f!l c:J tn
GJ GJ ~ G o m
us
WI
1:;:1
o
a
o
I!l
o
I!l C!l C!l

J: I [J [J IJ
Q
z .5755 o m o e
[;) 0 GOOD
- I! (J
(J
(J
[J
(J
[J
0
(;J 0 (J
0 (J

(J
0 0 CD
CD 0
0
0 0
UNDERSIZE
0

SPL.IINIES 0< ~ REJECT}

t:J GOOD . _ BY RING GAGE fiT

22 Gear Technology